362 ALASKAN GLACIER STUDIES
Billings Glacier is fairly clean, with no moraines of any size. Its lower end has a very moderate slope, and its front has apparently been in about the same position and condition in 1887, when it was first shown on a map by Applegate, in 1898, when some of the army officers of Glenn's party spoke of it as a "dead glacier," in 1909 when it was sketched by Grant and Higgins, and in 1910, when it was seen from the fiord by the National Geographic Society's expedition. No one seems to have gone near enough the glacier to study it in detail.
The Valley Train. The largest stream from Billings Glacier, Cabin Creek, flows southwest and south from the terminus of the glacier over an outwash gravel plain or valley train, receiving the stream from a small unnamed glacier on the west, and entering the fiord over a large delta, which has been built forward nearly a quarter of a mile into Passage Canal. The larger part of the outwash plain and delta is overgrown with willows, alders, and cottonwoods, showing not only that the glacier has not extended down to the fiord for many years, but that its melting does not supply as much water and detritus as formerly and that the building of the outwash of the valley train is now going on slowly; otherwise the streams would shift back and forth over the valley bottom and destroy the vegetation.
A Marginal Gorge. Another glacier stream from Billings Glacier enters the fiord a short distance to the northeast (g, Fig. 54) and is of special interest because between the glacier and the fiord it flows through a narrow, steep-sided rock gorge similar to but deeper than the abandoned gorges on the northern side of Toboggan Glacier. The gorge has been cut so nearly down to sea level that, through it, glimpses of the glacier surface may be had from the fiord. The stream seems to have established this course at a time of greater expansion of Billings Glacier when the ice was several hundred feet thicker than now, and extended high enough for a marginal stream to establish the present channel across the rock spur into which it is deeply incised. The glacier must have long remained in this more expanded condition to have permitted this gorge to be cut so nearly down to sea level that it is still occupied by a stream, although the ice tongue long ago retreated far enough to render this channel unnecessary. It is possible that this expanded stage occurred at the time when the glacier at the head of Passage Canal extended eastward, receiving Billings Glacier as a tributary, so that the broad valley now occupied by the main stream on the outwash gravel plain was entirely filled with ice, necessitating the cutting of this marginal channel.
Glaciers at the Head of Passage Canal—Portage Glacier Pass. There are three ice tongues at the western end of Passage Canal, of which the middle one, Portage Glacier, is the largest. It is of particular interest because it occupies a low pass across the Kenai Mountains which was habitually crossed by the Alaskan natives and by the Russians previous to the explorations of Vancouver and Whidbey in 1794, and probably later. The glacier now fills this pass but it has, nevertheless, been used as a highway by a number of United States army parties, and in 1898 by many prospectors. Their route, described by Mendenhall, led three quarters of a mile over an outwash gravel plain to the glacier, up Portage Glacier to the divide, and down the connecting through glacier on the western side, to an outwash gravel plain at a distance of about eight miles from the head of Turnagain Arm, a branch of Cook Inlet. The total length of this portage route is between 12 and 13